| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Environ. Engineering, Aalborg Univ., Sohngaardsholmsvej 57
b City and Environment Section, Aalborg Municipality, Vesterbro 14, DK-9000 Aalborg, Denmark
c Osozawa), Dep. of Regional Crops Science, Natl. Agric. Res. Center for Western Region, Senyu 1-3-1, Zentsuji, Kagawa, 765-8508 Japan
d Yamaguchi), Graduate School of Science and Engineering, Saitama University, 255 Shimo-okub, Saitama, 338-8570 Japan
e Soils and Biogeochemistry, Dep. of Land, Air, and Water Resources, Univ. of California, Davis, CA 95616
* Corresponding author (i5pm{at}civil.auc.dk)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
-10 to -15000 cm H2O. The SWC-dependent Buckingham-Burdine-Campbell (BBC) gas diffusivity model predicted DP data well within the same matric potential range for the 18 Andisols. The BBC model showed a minor but systematic underprediction of DP for three out of the four Gray-lowland soils, likely due to a blocky soil structure with internal fissures. A recent DP model that also takes into account macroporosity performed nearly as well as the BBC model. However, DP in the macropore region (air-filled pores >30 µm) was consistently underpredicted, likely due to high continuity of the macropore system in both Andisols and Gray-lowland soils. In agreement with previous model tests for 21 European soils (representing lower porosities and water retention properties), both SWC-dependent DP models gave better predictions for the 22 Japanese soils than soil-type independent models. Combining DP and SWC data, a so-called gas diffusion fingerprint (GDF) plot to describe soil aeration potential is proposed.
Abbreviations: BBC, Buckingham-Burdine-Campbell • GDF, gas diffusion fingerprint • IRC, incremental relative change in gas diffusivity • REV, representative elementary volume • RMSE, root mean square error • SWC, soil-water characteristic
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Volcanic ash soils are distributed across 0.84% of the earth's land surface (Leamy, 1984). For example, Andisols are found on 7.4 million ha in Southeast Asia (Fumoto and Sverdrup, 2000). The major areas with volcanic ash soils are located in the circum-Pacific region including Japan, Korea, New Zealand, the Philippines, and the western coast of North America (Bullard, 1984), but some are also found in Europe (Italy and France). The soils have a high potential for agricultural production because of their unique physical and chemical properties, including high water retention, good drainage, and high nutrient availability (Shoji et al., 1993). A comprehensive description of the genesis, properties, and utilization of volcanic ash soils is provided by Shoji et al. (1993).
Andisols show a wide variation in soil texture, but exact grain size distribution is difficult to determine, due to the typically high content of noncrystalline Al- and Fe-based materials that inhibit dispersion of mineral particles during analysis (Shoji et al., 1993). Thus, grain size distributions are of limited value in characterizing Andisols since sand, silt, and clay contents do not have as precise a meaning for Andisols as for soils consisting largely of crystalline minerals (Warkentin and Maeda, 1974). The physical characteristics of Andisols may therefore be more stringently defined by their soil-water characteristic curve (pore-size distribution) within zero to -15000 cm H2O of matric potential (Osozawa et al., 1990). Ito et al. (1991) observed a highly significant relationship between water retention at a matric potential of -15000 cm H2O and the content of both allophanic clays and humus. Drying to matric potentials below -15000 cm H2O makes volcanic ash soils less dispersible through irreversible aggregation (Kubota, 1976), a phenomenon not seen in other mineral soils (Iwata et al., 1995), and further illustrating the unique physical and chemical characteristics.
Andisols typically have a well-developed soil structure with a wide range of pore sizes that retain a large amount of water with varying matric potentials (Furuhata and Hayashi, 1980; Saigusa et al., 1987). Egashira et al. (1983) showed that water-stable aggregates in Japanese Andisols typically contain high amounts of clay-size particles (12–71%) and organic matter (3–35%). The high content of noncrystalline and organic materials affects the soil structure that typically exhibits either granular, angular blocky, or subangular blocky structure (Shoji et al., 1993). The high degree of soil structure normally found in Andisols and its dramatic effect on transport properties can be illustrated by the observation that light clay soils generally have saturated hydraulic conductivities <10-6 m s-1 while comparable (light clayey) Andisols have conductivities >10-4 m s-1 (Iwata et al., 1995).
Due to the often well developed and uniform soil structure with a small characteristic length (e.g., aggregate size), relatively small sample sizes can typically be used as representative elementary volume (REV) for undisturbed soil samples (Miyazaki, 1993). On the basis of measurements of water content and bulk density within different sample areas of an upland field location in Japan, and setting three criteria [(i) the REV must produce a small standard deviation between samples, (ii) the REV must be representative of the spatial structure, and (iii) the REV must provide an operational and convenient measurement method], Sato and Tokunaga (1976) recommended use of standard cylindrical 100-cm3 samples which cover a cross-sectional area of at least 20 cm2.
A dominant constituent of many Andisols is allophane, a noncrystaline, hydrous aluminosilicate. Allophane is a highly porous mineral with an outside diameter of 3 to 5 nm and wall thickness of 0.7 to 1.0 nm. Surface area measurements typically range from 580 to 1100 m2 g-1, depending on the measurement method (N2-Brunauer-Emmett-Teller or ethylene glycol monoethyl ether) (Shoji et al., 1993; Iwata et al., 1995). Particle density of allophanic Andisols ranges between 2.5 and 2.7 g cm-3, bulk density typically ranges between 0.4 and 0.8 g cm-3, and total porosity between 0.6 and 0.85 cm3 cm-3. The unusually high total porosity and content of micropores found in allophanic Andisols are primarily due to the intra- and interparticle pores of allophane.
Only information on Andisols of direct relevance to the present study has been introduced here. For more on morphology and mineralogical, chemical, and physical characteristics of Andisols, we refer to Shoji et al. (1993). In Shoji et al., a database of selected Andisols from Japan, Alaska, and Northwestern USA, prepared by the Soil Science Laboratory at Tohoku University, Japan, is described. For more on allophane characteristics and photomicrographs on Andisols illustrating soil structure, we refer to Iwata et al. (1995).
This study is based on SWC and gas diffusion coefficient data measured on 22 soils from Japan, including 18 Andisols. The measurements were done on undisturbed 100-cm3 samples in a broad, matric potential range between -10 (near water saturation) and -15000 cm H2O (wilting point). Most of the data have not previously been presented, and none of the data have been presented in English-language journals.
The main objectives of this study are (i) to test the ability of the Campbell SWC model to describe SWC data (and thus pore-size distributions) for the 22 soils, and (ii) to test the ability of recent gas diffusivity models (Moldrup et al., 1999, 2000), based on the Campbell SWC model, to predict the measured gas diffusivity data for the 22 undisturbed soils. The two SWC-dependent gas diffusivity models tested are also the two models recommended in the new Part 4 of Methods of Soil Analysis: Physical Methods (Dane and Topp, 2002; Section 4.3 on gas diffusivity) for predicting gas diffusivity in undisturbed soils. In this study, the models are validated against soils of widely different properties than before, illustrating the usefulness of the models for a wide range of soils. Additionally, we suggest a data presentation method that enables an easy and straightforward comparison of gas diffusivity and soil aeration potential in different soil types, in the form of a so-called GDF plot.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Gas diffusivity and water retention were measured on undisturbed soil samples at nine water potentials [pF = 1.0, 1.3, 1.5, 1.8, 2.0, 2.5, 3.0, 3.5, and 4.2, where pF = Log(-
; the matric potential in cm H2O)]. The sample area is characterized by humic and fine-textured Andisols with typically 30 to 50% clay, 25 to 40% silt, and 20 to 45% sand (predominantly fine sand). Tsumagoi 1 and 2 were sampled at the 0- to 5- and 20- to 25-cm depths of a cultivated field high in clay. Tsumagoi 3 to 5 were sampled at the 0- to 5- (loamy soil), 20- to 25- (clayey soil), and 44- to 49-cm (clayey soil) depth at a highly humic (typically 9 to 11% organic C) cultivated field. The main crop of both fields was cabbage. Tsumagoi 6 and 7 were sampled at the 25- to 30- and 76- to 81-cm depth at a noncultivated field. Tsumagoi 7 had a visibly different structure including small (mm-size) porous stones, a phenomenon known as a floating porous stone layer. The Tsumagoi sampling area is further described in Osozawa et al. (1994).
(ii) Five Andisols from Miura, Kanagawa Prefecture, Honshu (mainland Japan), labeled Miura 1 to 5.
Gas diffusivity and soil-water retention were measured on undisturbed samples at 8 water potentials (pF = 1.0, 1.5, 1.8, 2.0, 2.5, 3.0, 3.5, and 4.2). The sample area is characterized by light-clay Andisols. The main crop was Japanese radish (Raphanus sativus var. niger J. Kern.). Miura 1 to 3 were sampled at the 0- to 5-, 30- to 35-, and 50- to 55-cm depth at a cultivated field (deep plow treatment). Miura 4 and 5 were sampled at the 0- to 5- and 30- to 35-cm depth at a cultivated field where a soil layer exchange treatment had taken place 3 to 4 yr before sampling.
(iii) Six Andisols from Kumamoto prefecture in Kyushu (south Japan), labeled Kyushu 1 to 6.
Gas diffusivity and water retention were measured on undisturbed samples at six water potentials (pF = 1.0, 1.5, 2.0, 3.0, 3.5, and 4.2; data at pF 4.2 for Kyushu 4 were not available). The sample areas (grasslands) were characterized by humic and highly humic Andisols. Kyushu 1 to 3 were sampled at the 4- to 9-, 25- to 30-, and 45- to 50-cm depth at a highly humic field. Kyushu 4 to 6 were sampled at the 3- to 8-, 17- to 22-, and 45- to 50-cm depth at a humic field.
(iv) Four Gray-lowland soils from Kounosu, Saitama Prefecture, Honshu (1 soil, labeled KounosuGL) and Kumamoto Prefecture, Kuyshu (3 soils, labeled KyuGL 1-3).
Gas diffusivity and water retention were measured on undisturbed samples at six water potentials (pF = 1.5, 2.0, 2.5, 3.0, 3.5, and 4.2 for KounosuGL; pF = 1.0, 1.5, 2.0, 3.0, 3.5, and 4.2 for KyuGL 1 to 3). The soils were from paddy fields and the crop was rice. KounosuGL is a high-clay soil sampled at 0 to 5 cm. KyuGL 1 and 2 are mineral soils sampled at the 5- to 10- and 22- to 27-cm soil depth. KyuGL 3 is a buried organic layer (kuroniga) located at the 78- to 83-cm depth at the same field as KyuGL 1 and 2.
Measurement Methods
Soil-water retention was measured by the method of Klute (1986). The intact soil cores (100-cm3 sample volume) were saturated in sand boxes and subsequently drained to the desired matric potentials () using a hanging water column ( 
-30 cm H2O; pF 1.5) or a pressure plate apparatus ( < -30 cm H2O). After each drainage step, gas diffusivity was measured on the samples. The experimental setup (diffusion chamber) was first suggested by Taylor (1949) and further developed as described by Schjønning et al. (1985) and Osozawa (1987). Soil gas diffusion was measured with oxygen as the experimental gas at 20°C. The same measurement method was used by Kruse et al. (1996) and Moldrup et al. (2000). Oxygen consumption could be considered negligible during the short periods needed to measure the diffusion coefficient at each matric potential (Moldrup et al., 2000). The diffusion coefficient was calculated by the method of Currie (1960).
Soil-Water Characteristic Model
The Campbell (1974) SWC model is
 | [1] |
is the matric potential (cm H2O), e is the matric potential at air-entry (cm H2O), 
is the volumetric water content (m3 m-3), s is the water content at saturation (m3 m-3), and b is the Campbell pore-size distribution parameter (b > 0), corresponding to the slope of the SWC curve in a Log()–Log() coordinate system.
Gas Diffusivity Models
The most frequently used soil-type-independent gas diffusivity models are the classical equations suggested by Penman (1940) and Millington and Quirk (1960)(1961). The Penman model is
 | [2] |
is the volumetric soil-air content (air-filled porosity; m3 soil air m-3 soil).
The Millington-Quirk (1960) and (1961) models also include total porosity (
, m3 m-3), and are
 | [3] |
 | [4] |
Moldrup et al. (1999) suggested the so-called BBC model,
 | [5] |
2 corresponds to gas diffusivity in completely dry soil, as suggested by Buckingham (1904) and the term 2+3/b is an analogue to the Burdine (1953)–Campbell (1974) tortuosity model for describing unsaturated hydraulic conductivity (Moldrup et al., 1996). Equation [5] was developed based on DP and SWC data for 20 undisturbed soils with b values ranging from 2 to 11, and successfully tested against data for 6 undisturbed soils with b values ranging from 4.6 to 14. Total porosities for the 26 soils were typically <0.6 m3 m-3 and organic C contents typically <2.5%.
Moldrup et al. (2000) found a highly significant correlation (coefficient of regression, r2 = 0.97) between air-filled porosity at -100 cm H2O of matric potential (100; corresponding to the volume of soil pores with equivalent pore diameter >30 µm) and gas diffusivity at -100 cm H2O (DP,100),
 | [6] |
Equation [6] described data well for 126 undisturbed soils representing a wide range of soil texture and soil management (Moldrup et al., 2000). The good agreement between Eq. [6] and measurements is illustrated for 144 undisturbed European soils in Moldrup et al. (2001), resulting in a relatively narrow 95% prediction interval. It is noted that for most of the 144 soils, only measurements at -100 cm H2O of matric potential were available, and thus the data could not be used for overall DP() model tests.
Using Eq. [6] as reference-point gas diffusivity in combination with the Burdine-Campbell tortuosity model yields the following model for DP() (Moldrup et al., 2000),
 | [7] |
To compare gas diffusivity models, the root mean square error (RMSE) of prediction was used for best overall fit compared with measured data,
 | [8] |
 | [9] |
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
) and the () data (Fig. 1), suggesting a close relationship between the gas diffusivity and SWC relationships.
|
|
e were generally around or <-10 cm H2O and are not provided for each soil as this parameter is not included in the SWC-dependent gas diffusivity models. Since the simple two-parameter Campbell model provided near-perfect fits to the measured SWC data, multi-parameter SWC models like the van Genuchten (1980) model were not considered in this study.
The often significant difference between the total porosity and the volumetric soil-water content at pF 1 (Table 1) implies a relatively high content of pores > 300 µm equivalent pore diameter, suggesting a high degree of structure and aggregation. Generally, the soils show extremely high water retention as illustrated by the steep SWC relations in Fig. 2. Except for KyuGL 2 with a total porosity of = 0.49 cm3 cm-3, the total porosity ranges between 0.60 (KyuGL 1) and 0.87 for the kuroniga buried organic layer (KyuGL 3). For the 18 Andisols, ranges between 0.69 and 0.82 (Table 1).
Both SWC-dependent gas diffusivity models predict the measured gas diffusivities in the undisturbed Andisols (Fig. 3) and the undisturbed Gray-lowland soils (Fig. 4) well. Figure 3 shows data and model predictions for 15 of the 18 Andisols. The remaining three Andisols (Tsumagoi 4 and 6, Kyushu 6) have b values between 10 and 15 and show similar behavior as the soils in Fig. 3.
|
|
) curveshape is different from the other Tsumagoi soils and implies a dual-porosity pore system with different gas diffusion properties below and above pF 2. The five Miura soils show very similar water retention and gas diffusivity properties, and no differences in physical properties related to the different soil cultivation techniques used at Miura 1 to 3 and 4 to 5, respectively, were apparent. Contrary to this observation, the six Kyushu Andisols showed extremely different water retention and gas diffusivity properties ranging from Kyushu 2 with very little drainage occuring between pF 1.0 and 4.2 to Kyushu 4 with high drainage ability and high gas diffusivities. It is encouraging that both SWC-dependent models satisfactorily predict gas diffusivity when tested against soils representing a wide range of pore-size distributions, extremely high porosities and, in some cases, very high organic matter contents (Fig. 3).
For the three higher mineral Gray-lowland soils, a minor but systematic underprediction by both SWC-dependent models was evident (Fig. 4a,b,c). The likely explanation is that rice plant roots in such paddy fields create an angular blocky soil structure with small fissures and thus a very high continuity of the large pores, as illustrated by Osozawa (1998)(p. 18) and Iwata et al. (1995)(p. 278). This likely creates a dual-porosity pore system, explaining the bend of the DP() curve around pF2 (at the 100 value). For the kuroniga buried organic layer, both the 100–dependent and, especially, the BBC model well predict the measured gas diffusivities (Fig. 4d) hereby representing the first test of the SWC-dependent models against an almost completely organic porous medium.
At first glance, the many gas diffusivity relations in Fig. 3 and 4 appear similar. A closer look into the data, however, reveals the soil-type dependency of DP(). Figure 5
shows gas diffusivity data for two soils, one representing the lower range of b values (b = 8.5) and one the higher range (b = 28). The more clayey/organic soil (Kyushu 3) is highly structured and exhibits larger gas diffusivities at the same soil-air content, as compared with the less clayey soil (Tsumagoi 3).
|
) models, since they will all pass through the origin (DP/D0 = 0 for = 0). To evaluate this further, we applied the model by Troeh et al. (1982),
 | [10] |
0.02 m3 m-3) and surprisingly similar for the two very different soils. In comparison, for the KyuGL 3 soil with similar pore-size distribution (same b value) and near-identical DP data in the low DP range as the Kyushu 3 soil, the fitted u value was three times higher (0.06 m3 m-3). Generally, the value of u varied between almost zero and 0.128 cm3 cm-3 (Miura 4) for the 22 soils. The value of u did not correlate with porosity or SWC properties, and the model fit was very sensitive to even small variations in measured DP data when extrapolating the model outside the range where measurements were available.
In line with this, Petersen et al. (1994) fitted a u value of 0.12 m3 m-3 (v = 1.23) for repacked Yolo light clay using DP measurements at > 0.12 m3 m-3 while Sallam et al. (1984), for the same soil packed at the same bulk density and using the same gaseous tracer (Freon), had measured significant DP values for as low as 0.05 m3 m-3. In conclusion, the Troeh et al. model (Eq. [10]) proved useful for accurately fitting measured DP() data within the interval where measurements were available (Fig. 5; Petersen et al., 1994). The model is therefore very applicable in gas diffusion and fate models where DP data within the region of interest are measured, as illustrated by Petersen et al. (1996). However, the Troeh et al. model could not be used to verify the existence (or evaluate the magnitude) of a threshold soil-air content nor be used as a predictive gas diffusivity model, since no relationship between the fitting parameters (u and v) and other physical properties was evident.
Both SWC-dependent gas diffusivity models (Eq. [5] and [7]) did not show any general tendency to overpredict (or underpredict) measured DP/D0 values in the low DP/D0 range (Fig. 6d and 6e)
. Therefore, the inclusion of a threshold soil-air content does not appear necessary for accurately predicting gas diffusivity even in the very low range (wet soil). The same was seen for 21 European soils representing lower porosities and soil-water retention properties (Moldrup et al., 2000). We do not discount the existence of isolated air pockets in the soil, but the phenomenon seems to take place a very low soil-air contents (Schjønning et al., 2002) where the proposed SWC-dependent models in practice will yield predicted DP values very close to zero.
|
) model concept seems promising for describing the observed soil-type effects on gas diffusivity, and soil-type-dependent models appear necessary for realistically predicting gas diffusivity in undisturbed soils. This was also evident when testing both soil-type independent (Eq. [2]–[4]) and soil-type-dependent (Eq. [5]–[7]) gas diffusivity models against the measured data for the 22 soils (Fig. 6). Looking at the two most frequently used soil-type independent gas diffusivity models, the Penman (1940) model largely overestimates and the Millington and Quirk (1961) model largely underestimates the measured DP data in Fig. 6a,c. This is also evident for the two soils depicted in Fig. 5. The less known and infrequently used Millington and Quirk (1960) model performed the best among the soil-type independent models, Fig. 6b and Table 2.
|
The soil-type-dependent (SWC) models generally performed better than the soil-type independent models (Figs. 6d and e). The BBC model, Eq. [5], performed the best among all five models tested, both for Andisols and Gray-lowland soils (Fig. 6d). The macroporosity-dependent (100) model, Eq. [7], in combination with Eq. [6], performed almost as well (Fig. 6e) but systematically underestimated the higher diffusivities (corresponding to the diffusivities at higher soil-air contents). This resulted in a relatively high, negative bias (Table 2). In contrast, the 100–dependent model yielded zero bias for the European soil data, that is, no general under- or overestimation of measured DP data.
The reason for the different model performance in regard to the European and the Japanese soils is found in Eq. [6]; that is, when predicting DP/D0 in the macroporous pore space (with >30-µm equivalent pore diameter). When separately testing Eq. [6] against the measured DP,100/D0 values, a minor but systematic underprediction is seen (Fig. 7) . The overall performance of Eq. [6] must be considered good since the measured data for most of the 22 soils are within the 95% prediction interval found by Moldrup et al. (2001) when testing the model against data for 144 undisturbed European soils. However, the measured DP,100/D0 values for the Andisols and Gray-lowland soils are in most cases above the predicted curve (Eq. [5]) suggesting a less tortuous macropore system in the well structured Andisols and Gray-lowland soils compared with the 21 European soils.
|
) values for both Andisols and Gray-lowland soils (Fig. 6f) with nearly zero bias (Table 2). Thus the Burdine-Campbell tortuosity functions (2 + 3/b) seems to accurately describe the change in gas diffusivity with soil-air content, but the reference-point model for gas diffusivity in the macropore volume (Eq. [6]) does not capture the larger pore continuity in the well-structured and highly-porous Andisols and Gray-lowland soils, as compared with normal mineral soils such as the 21 European soils. On the basis of the above model tests, recommendations can be made. If a measured value of DP,100 is available, Eq. [7] appears to be the superior model choice. If not, the BBC model, Eq. [5], provided robust gas diffusivity predictions across soil types and porosities with minimum input parameter requirement. Thus, the BBC model is recommended instead of the traditionally used soil-type independent model in gas transport-reaction models applied to undisturbed soils.
So far, this study has focused on generalized predictive models for gas diffusivity, taking into account soil type and structure. However, as the data for the Japanese soils are more detailed (measured at more pF values) than previous gas diffusion studies, the data allowed for a more thorough analysis of differences in soil aeration potential between soils types. If detailed DP and SWC data are available, we propose to make a 
(DP/D0)/ vs. pF plot, where the (DP/D0) and values are calculated as the increments obtained between two consecutive pF values. This will illustrate the incremental relative change in gas diffusivity (labeled IRC, typically 0 < IRC < 1) within each pF interval and, thus, within each pore-size interval. This should originally be plotted as a histogram but in order to facilitate an easy-to-read plot even when data for several soils are shown together, the corresponding pF value for each IRC is suggested to be taken as the mean of the two consecutive pF values. Thus, each IRC value is calculated as
 | [11] |
This type of plot, labeled a GDF plot, is shown in Fig. 8 for six selected soils. For the three less clayey and less humic Andisols from Miura (2, 4, and 5), a rapid increase in IRC is already observed from pF 1 (-10 cm H2O of soil matric potential) and high IRC values are obtained throughout the entire soil matric potential interval. This suggests an optimal soil aeration potential and, accordingly, the Miura soils are known to be without crop disease problems related to poor soil aeration (Osozawa, 1998). High IRC values (close to 1) are seen around pF 2, suggesting a very high continuity of the air-filled pores drained at this soil matric potential. No significant differences in the GDF plots for the three Miura soils were found (Fig. 8) in spite of the different soil cultivation methods applied to Miura 2 as compared with Miura 4 and 5. Thus, the two cultivation methods apparently did not cause significant differences in soil structure and pore network organization, likely because the exchange layer treatment on Miura 4 and 5 was carried out 3 to 4 yr before the sampling, allowing the original soil structure to be mostly reestablished. Such a partial reestablishment of soil structure after a dramatic soil disturbance has earlier been shown by Schjønning et al. (1999) for loamy and clayey soils 1.5 years after soil disturbance.
|
pF 2.5 (-300 cm H2O) and the IRC values are low throughout the remaining pF interval (Fig. 8). This suggests poor soil aeration potential, consistent with the observation that the cabbage crops on these soils were suffering from clubroot disease (Osozawa et al., 1994). In contrast, the GDF for the less clayey Tsumagoi 3 soil (not shown) implied good soil aeration potential and was very similar to the GDFs for the Miura soils. Interestingly, the kuroniga buried organic layer (KyuGL 3) showed an almost linear increase in IRC with pF (Fig. 8), suggesting a very different pore network organization compared with the other soils. The GDF plot, based on combining detailed SWC and gas diffusivity measurements, seems useful to compare aeration potentials for different soil types and soil management methods, and can likely also assist in evaluating changes in pore network organization caused by soil management. |
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The BBC SWC-dependent gas diffusivity model (Moldrup et al., 1999) well predicted gas diffusion coefficients for the 22 soils in the same matric potential interval. A minor but systematic underprediction for three paddy field soils can be explained by the angular blocky soil structure.
The recent SWC-dependent gas diffusivity model by Moldrup et al. (2000), also based on the Burdine-Campbell tortuosity model but taking into account soil macroporosity, gave almost as good predictions. However, the model systematically underpredicted gas diffusivity at -100 cm H2O of matric potential, likely due to the well developed soil structure of the Andisols and Gray-lowland soils causing higher macropore continuity compared with normal mineral soils.
Correcting the gas diffusivity model of Moldrup et al. (2000) for the higher macropore continuity, by using the measured gas diffusivity at -100 cm H2O as a reference point, enabled an accurate description of the gas diffusivity data across the entire matric potential interval. Hence, the Burdine-Campbell tortuosity model seems generally valid to describe changes in gas diffusivity with air-filled porosity in undisturbed soil across soil types and porosities.
Both for the 22 Japanese soils and for 21 European soils representing smaller Campbell b values, lower organic C contents and lower total porosities, the widely used soil-type independent gas diffusivity models by Penman and Millington & Quirk performed poorly compared with the SWC-dependent models. This study emphasizes that only soil-type-dependent models can give realistic predictions of gas diffusivity in undisturbed soil across soil types. Both SWC-dependent gas diffusivity models recommended in the new Part 4 of Methods of Soil Analysis (Dane and Topp, 2002) proved to give reliable and accurate predictions, even when tested for a much wider range of soil types than considered before. As the BBC model has the least input parameter requirement and gives robust gas diffusivity predictions across soil types and porosities, this model is recommended for use in gas transport-reaction models applied to undisturbed soil systems.
A GDF plot, based on combining detailed gas diffusivity and water retention data, was proposed and may be used to easily distinguish between soils with good and poor aeration potential and could assist in evaluating differences in pore network organization.
|
ACKNOWLEDGMENTS |
|---|
Received for publication June 29, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. C. Resurreccion, P. Moldrup, K. Kawamoto, S. Yoshikawa, D. E. Rolston, and T. Komatsu Variable Pore Connectivity Factor Model for Gas Diffusivity in Unsaturated, Aggregated Soil Vadose Zone J., April 14, 2008; 7(2): 397 - 405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Hamamoto, T. Tokida, T. Miyazaki, and M. Mizoguchi Dense Gas Flow in Volcanic Ash Soil: Effect of Pore Structure on Density-Driven Flow Soil Sci. Soc. Am. J., February 15, 2008; 72(2): 480 - 486. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. G. Poulsen, P. Moldrup, S. Yoshikawa, and T. Komatsu Bimodal Probability Law Model for Unified Description of Water Retention, Air and Water Permeability, and Gas Diffusivity in Variably Saturated Soil Vadose Zone J., October 3, 2006; 5(4): 1119 - 1128. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. W. Bolton, R. A. Berner, and S. T. Petsch The Weathering of Sedimentary Organic Matter as a Control on Atmospheric O2: II. Theoretical Modeling Am J Sci, October 1, 2006; 306(8): 575 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Liu, B. Li, K. Hu, and M. Th. van Genuchten Simulating the Gas Diffusion Coefficient in Macropore Network Images: Influence of Soil Pore Morphology Soil Sci. Soc. Am. J., June 21, 2006; 70(4): 1252 - 1261. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ritter, R. Munoz-Carpena, C. M. Regalado, M. Javaux, and M. Vanclooster Using TDR and Inverse Modeling to Characterize Solute Transport in a Layered Agricultural Volcanic Soil Vadose Zone J., May 12, 2005; 4(2): 300 - 309. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Moldrup, T. Olesen, S. Yoshikawa, T. Komatsu, and D. E. Rolston Three-Porosity Model for Predicting the Gas Diffusion Coefficient in Undisturbed Soil Soil Sci. Soc. Am. J., May 1, 2004; 68(3): 750 - 759. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Caron and N. V. Nkongolo Assessing Gas Diffusion Coefficients in Growing Media from in situ Water Flow and Storage Measurements Vadose Zone J., February 1, 2004; 3(1): 300 - 311. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Moldrup, S. Yoshikawa, T. Olesen, T. Komatsu, and D. E. Rolston Air Permeability in Undisturbed Volcanic Ash Soils: Predictive Model Test and Soil Structure Fingerprint Soil Sci. Soc. Am. J., January 1, 2003; 67(1): 32 - 40. [Abstract] |
; the soil matric potential in cm H2O), and Dp/D0 is the relative soil-gas diffusion coefficient.
